While TV-interviews are recorded for an already written documentary film script, interviews for the Ukrainian oral history archive of war testimonies are recorded without any expectations. When creating our archive, we do not know what research question of linguists, philosophers or historians will be later. We know that the archive is not going public, that´s the deal.
But how does one study interviews in the archives, if there are not 10 interviews, but 100? And what if there are 1000 of them? Where do you get enough time to listen? This is where artificial intelligence comes to the rescue, technologies for “distanced reading”, “text mining” and “natural language processing”. The talk focuses on the technologies which are helping to get an overview over the interview collecting by indexing, keywording and transcription.
Afficher cette publication dans notre dépôt institutionnel (orbi.lu).